In one line
Clinical ethics in obstetrics and gynaecology is the disciplined balancing of four obligations — respect for autonomy, beneficence, non-maleficence and justice — against the law of the place you practise; in South Africa that means a competent adult's informed refusal is binding even when it risks the fetus (which has no separate legal personhood antenatally), valid consent is the reasonable-patient material-risk standard of Castell and Montgomery rather than whatever doctors customarily disclose, and the defensible plan is the one you can justify on a principle and a statute, not on a personal moral verdict.
This chapter assumes the day-to-day mechanics of taking consent and works one level up — the dilemmas where principles and the law pull against each other. The groundwork sits in the Intermediate chapter on informed consent; revise that for the elements and the consent form, and spend the effort here on capacity assessment, refusal, the maternal–fetal interface and conscientious objection.
Mechanism & pathophysiology
Ethics has no histology, but it has a structure, and reasoning fails when the structure is skipped. The dominant working model in clinical medicine is the principlism of Beauchamp and Childress: four prima facie principles, each binding unless it conflicts with another, none lexically superior, and the work is the balancing when two collide.
- Respect for autonomy — the patient's right to make her own decision about her own body, provided she has capacity and the decision is voluntary and informed. In O&G this is the strongest principle, because almost every decision is about a woman's own reproductive body and the historical paternalism here was severe.
- Beneficence — acting in the patient's interest, providing net benefit.
- Non-maleficence — primum non nocere; not inflicting net harm. Beneficence and non-maleficence are distinct: a Caesarean both benefits and harms, and weighing the two is the clinical act.
- Justice — fairness in the distribution of benefits, risks and resources. In a constrained state system this is not abstract: every theatre slot, ICU bed and dose of an expensive drug allocated to one woman is withheld from another.
A principle is prima facie, not absolute — it holds unless a weightier obligation overrides it in the specific case. Autonomy usually trumps a doctor's view of beneficence, which is why a competent refusal stands; but autonomy does not compel a doctor to provide a futile or harmful intervention on demand, because autonomy is a liberty (the right to refuse) more than a claim (the right to command any treatment).
The principles sit on older theory worth being able to name. Consequentialism (the right act is the one with the best outcomes) underlies cost-effectiveness and resource arguments. Deontology (some acts are right or wrong regardless of outcome — duties, rules, rights) underlies "you may not deceive the patient even for her own good". Virtue and care ethics ask what a person of good character, attentive to the particular relationship, would do — useful precisely where rules run out, as in breaking bad news or sitting with a refusing woman. No single theory resolves real cases; the principles are a shared vocabulary that lets a multidisciplinary team reason together.
When a case resists the principles, the structured case-analysis method of Jonsen, Siegler and Winslade — the four quadrants — organises it: medical indications (the facts: diagnosis, prognosis, what each option offers), patient preferences (her competent, informed wishes — autonomy made concrete), quality of life (the outcome she can expect, judged from her perspective not the team's), and contextual features (family, resources, law, institutional and social factors). Filling the four boxes forces the facts to the surface before the moral argument and stops a team arguing values when it is actually disagreeing about prognosis.
Assessment
The assessment in clinical ethics is the assessment of consent and capacity, because a plan that is technically correct but inadequately consented is unlawful and indefensible.
Valid consent has three elements, and all three must hold:
- Capacity — the patient can make this decision now.
- Disclosure — she has been given the information a reasonable patient in her position would want.
- Voluntariness — the decision is her own, free of coercion or undue influence (a controlling partner, a domineering family, a clinician's pressure).
Capacity is decision-specific, time-specific and functional, never a global label. A woman is not "competent" or "incompetent" in the abstract; she has capacity for a particular decision at a particular moment. The functional test asks whether she can understand the relevant information, retain it long enough to use it, weigh it in the balance, and communicate a choice. Failing any one defeats capacity for that decision. Capacity is presumed in an adult; the burden is on the clinician who would override it. Crucially, an unwise decision is not in itself evidence of incapacity — a competent adult may refuse life-saving treatment for reasons others find irrational or for no stated reason at all. Fluctuating capacity (sepsis, eclampsia, opioid analgesia, acute pain, extreme distress) demands reassessment at the moment of decision, not reliance on an earlier finding.
Disclosure is governed by the reasonable-patient material-risk standard, not by professional custom. South African law settled this in Castell v De Greef and the United Kingdom in Montgomery v Lanarkshire: a risk is material if a reasonable person in the patient's position would attach significance to it, or if the doctor is or should be aware that this particular patient would. The old "reasonable doctor" (Bolam/Sidaway) test — disclose what a responsible body of doctors would disclose — no longer governs what risks must be discussed. The practical consequence in O&G is concrete: you cannot omit the small-but-serious or the common-but-troubling risk (anal sphincter injury at instrumental delivery, hysterectomy as a possible end-point of a PPH, the failure rate of sterilisation, venous thromboembolism with combined contraception) on the grounds that "we don't usually mention it". Disclosure includes the diagnosis, the nature and purpose of the proposed intervention, the material risks and benefits, the reasonable alternatives (including doing nothing), and an honest account of uncertainty.
In South Africa the statutory frame is the National Health Act 61 of 2003: s7 prohibits providing a health service without the user's informed consent and requires the provider to take all reasonable steps to obtain it, while s6 obliges the provider to inform the user of her health status in a language and manner she understands. The Act also states that a user capable of understanding must be informed even where she lacks the legal capacity to consent — the ethical core of involving patients in decisions about themselves.
Voluntariness in O&G is not a formality. A consent signed under a partner's gaze, or by a young woman whose family has decided for her, or in the coercive context of intimate-partner control, may be invalid. Where coercion is suspected, the assessment is to see the woman alone.
Documentation is part of the consent, not a separate clerical act. The defensible record is a note of the conversation — what was explained, the material risks named, the alternatives discussed, the patient's questions and her decision — not merely a signed form. A signature on a form with no record of disclosure is weak evidence; a contemporaneous note of a genuine discussion is strong evidence even without a perfect form. Consent is a continuing dialogue, in the HPCSA's phrasing, revisited as the clinical picture changes — the consent taken at booking for a vaginal birth is not consent for an intrapartum Caesarean unless that conversation has been had.